Microchemical Journal 154 (2020) 104655 Contents lists available at ScienceDirect Microchemical Journal journal homepage: www.elsevier.com/locate/microc Application of statistical analyses for lapis lazuli stone provenance T determination by XRL and XRF ⁎ Miriam Saleha, Letizia Bonizzonia, , Jacopo Orsillia, Sabrina Samelaa, Marco Garganoa, Salvatore Galloa,b, Anna Gallic,d a Dipartimento di Fisica “Aldo Pontremoli”, Università degli Studi di Milano, via Celoria 16, 20133 Milano, Italy b Istituto Nazionale di Fisica Nucleare (INFN) sezione di Milano, via G. Celoria 16, 20133 Milano, Italy c Dipartimento di Scienza dei Materiali, Università degli Studi di Milano-Bicocca, via R. Cozzi 55, 20125 Milano, Italy d CNR-IBFM, via F.lli Cervi 93, 20090 Segrate, Italy ARTICLE INFO ABSTRACT Keywords: Lapis lazuli use stretches back more than 6500 years; ancient civilizations of Mesopotamia, Egypt, China, Greece, Radio-luminescence and Rome treasured and prized it. Afghanistan has been the oldest source for this stone, while Chile, Canada, X-Ray Fluorescence Russia and a few other countries have been reported as sources for raw material in more recent times; the rarity X-Ray Luminescence of historical mines surely represents a positive aspect for the provenance clue of artefacts. Lapis lazuli is a rock Lapis lazuli provenance consisting mainly of lazurite, to which it owes the blue colour, calcite and pyrite. Other constituents may be MV analysis present, related to the different mines. In the present work, we apply the principles of Radio-luminescence (RL) SAM analysis exploiting as radiation source the X-ray tube of a portable commercial X-Ray Fluorescence spectrometer; in this way, X-Ray Fluorescence spectra (XRF) can be simultaneously acquired to have a larger set of data. To highlight the instrumental experimental differences, we refer to the portable set up as X-Ray Luminescence (XRL),as suggested by recent literature. We thus looked for the possibility of applying a wieldy, low cost and non-de- structive method that could fit also to precious objects, based on the join use of XRF and XRL. We performed analyses on raw lapis lazuli stones from five different provenances, both historical and modern, and on foursets of unknown origin carved polished stones, to test our methods on real artefacts. We focalised on a limited number of samples to concentrate on the statistical treatment of spectra obtained, so to get a synergic response of the two applied techniques. We were able to obtain a clear distinction for the different classified provenances and could speculate those of unknown samples. 1. Introduction Luminescence) with the aim to go “towards a portable X-ray lumines- cence instrument for applications in the Cultural Heritage field” [7]. Lapis lazuli is one of the first rock used by ancient civilizations: Lapis sources are rare, due to the peculiar geological conditions artefacts have been found in prehistoric tombs in Asia, Africa, and required for its forming [8]; notwithstanding, only few studies are fo- Europe. It is composed by different minerals; the principal one, which cused on the determination of classification methods [9] or give an gives the blue colour, is lazurite. One of the mineral phases most exhaustive characterization of old sources. This lack of systematic stu- commonly associated to it is diopside, while other ancillary minerals dies can be partially ascribed to the confusion about old sources as can be calcite, pyrite, feldspars, wollastonite. Their presence may vary explained hereafter. Throughout the recorded use of lapis, the principal and may be related to lapis lazuli provenance [1]: in fact, different source has been in the Badakshan district of Afghanistan, which has mineral phases in stones contains typical trace elements, that can be been working for approximately six thousand years and has been detected by means of elemental analyses, such as PIXE (Particle Induced mentioned frequently in historical documents. A Siberian deposit is X Emission) [2] or XRF (X-Ray Fluorescence) [3]. Exploiting the lu- located for sure at the southern end of Lake Baikal. The only other minescence properties of minerals, also Ion Luminescence (IL) or Radio important locality is the Ovalle Cordillera (Andes Mountains), Co- Luminescence (RL) can be used for provenance studies [4–6]. Recent quimbo Province, Chile. Other localities of minor importance include papers have also proposed the application of XRL (X-Ray sites in Myanmar, Colorado, and California. ⁎ Corresponding author at: Dipartimento di Fisica, Università degli Studi di Milano, via Celoria 16, 20161 Milano, Italy. E-mail address: [email protected] (L. Bonizzoni). https://doi.org/10.1016/j.microc.2020.104655 Received 29 August 2019; Received in revised form 13 January 2020; Accepted 16 January 2020 Available online 18 January 2020 0026-265X/ © 2020 Elsevier B.V. All rights reserved. M. Saleh, et al. Microchemical Journal 154 (2020) 104655 In modern jewellery trade, various qualities of lapis lazuli are 2.2.2. Unknown origin samples named after the region of origin, namely Chilean Lapis (the least va- Some unknown stones have also been considered, to better test the luable type; it contains numerous white calcite inclusions and is often classification capability of the proposed methods on real precious ob- tinged or spotted with green), Russian, or Siberian, Lapis (it contains jects, as most of the techniques proposed in literature are not applicable pyrite, and is usually of good quality) and Persian Lapis, which is ac- in this peculiar case. The stones are mounted on pieces of jewellery tually from Afghanistan (the finest, with little or no pyrite and nowhite from private collections and have not been disassembled. For these calcite veining). Commercial names confirm the confusion about source samples, only the purchase area is known; preliminary reflectance locations: this also contributes to the difficulty in finding a correlation spectroscopy measurements have been conducted to confirm the geo- between mineral constituents and provenance. Most of the historical logical nature of the stone. The four collections are set out in Table 1. sources are located in quite inaccessible sites and it is not difficult to Collection 1 is a complete parure, but single pieces show some understand the dearth of eyewitness reports of activities at the mines, differences in mounting. Namely, the ring has geometrical decorations especially in old times. Indeed, many museums collections still show a instead of floral decorated filigree. The earrings and brooch mountings lack of information about the exact provenance of stones [9]. Scientific have different Cu/Au and Pb/Au ratio, as verified by XRF, evenifthe literature also reports analyses on Lapis from Pakistan, Iran and Persia decoration of the metallic parts is very similar. Thus, it is possible that [10–12]. It worth noting that, from the geological point of view, Persia the set was originally not complete, and the stones themselves could be overlaps with nowadays Iran and a part of Iraq. Old manuscripts from different: in fact, also appearance of stones shows some differences in XIII and XIV century, among which the diary of a Mongolian Empire the shades of blue and in the presence of gold colour speckles, very emissary, report local mines, but they might be only a small quarry evident for instance in the ring stone. Regarding Collection 3, it has exhausted in the XIV century [13]. been produced and acquired in Ischia, an island near Naples, a region In this complex landscape, we looked for the possibility of applying notorious for the carving of semi-precious stones and coral. Anyhow, a a portable, low cost and non-destructive method that could fit also to local origin (Vesuvius) for the stones can be excluded without any precious objects, based on the join use of XRF and XRL. We developed doubt, since no commercial exploitation of eventual local lapis lazuli then a statistical treatment of spectra obtained, so to get a synergic quarry is attested. These sets of samples are intended to verify if our response of the two applied techniques. For this reason, we focalised on reference groups are able to classify unknown provenance stones. a limited number of samples, comparing results obtained by different data elaboration. 2.2. Analytical techniques 2. Materials and methods Aiming to obtain a methodological approach exploiting a relatively low cost and portable instrumentation, a prototype was assembled, 2.1. Lapis lazuli samples starting from a commercial portable XRF spectrometer and a lumines- cence detector. We applied the principles of Radio-luminescence (lu- Five different provenance lapis lazuli samples were selected totest minescence induced by X-Rays) exploiting as radiation source the X-ray the applicability of the proposed method, considering both polished tube of the portable commercial X-Ray Fluorescence spectrometer de- and unpolished stones to verify whether the texture could interfere with scribed in Section 2.2.1; in this way, X-Ray Fluorescence spectra (XRF) classification. More than one point on each sample was considered, can be simultaneously acquired to have a larger set of data. To highlight when possible, to take into account the intrinsic non-homogeneity of the instrumental experimental differences, we refer to the portable set the rocks. up as X-Ray Luminescence (XRL) and to traditional laboratory one as RL (Radio-luminescence). 2.2.1. Known origin samples The described configuration